Wan image model
Image-to-image model designed for controlled restyling, product cleanup, and source-image refinement.
Wan is a job-specific creative surface
This page is designed around controlled image refinement: a quick generator first, followed by examples, settings, comparison guidance, and trust notes that explain why this model is the right starting point.
Image-to-image cleanup and controlled restyling.
Not the best first choice when no source image exists.
Use Nano Banana Pro for text-to-image campaigns or Qwen Image for brand-system continuity.
What Wan is built to create
Product Cleanup
Removes clutter while keeping product identity.
Style Transfer
Applies a new finish without rebuilding the whole image.
Poster Cleanup
Improves a draft poster into a cleaner visual base.
How to start with Wan
Upload the source image and list the preserved elements first.
Separate cleanup instructions from style instructions.
Use subtle or balanced strength for product identity.
Compare variants in the creation library before using them in campaigns.
Edit discipline
Wan is strongest when the user can name exactly what should stay and exactly what should change.
Useful constraints
Preserve composition, preserve product label, remove background clutter, improve light, and keep color family are practical constraints.
Where it fits
Use it for polishing existing assets, cleaning product shots, reworking poster drafts, and producing controlled visual variants.
Trust details that help users create with context
The page explains the recommended mode, starting settings, and where a model may need additional iteration before production use.
Review prompts, people, brands, text, and usage rights before publishing generated assets in campaigns or client work.
Model guidance is organized by realistic creative jobs such as product motion, story scenes, catalog images, posters, and brand extensions.
Run a first Wan pass instead of only reading docs
This model page helps you decide fit. Once the job is clear, carry the prompt into Studio and generate the first pass.
What is Wan?
Wan is an Imaveo image model designed for image-to-image cleanup and controlled restyling.
The page is organized around controlled image refinement, with a quick Studio handoff, example gallery, comparison notes, and trust guidance.
How to use Wan on Imaveo
- Step 1Upload the source image and list the preserved elements first.
- Step 2Separate cleanup instructions from style instructions.
- Step 3Use subtle or balanced strength for product identity.
- Step 4Compare variants in the creation library before using them in campaigns.